The intravacuum mirrors used in the systems of optical plasma diagnostics on fusion installations are subject to contamination and require periodic cleaning. It is proposed to clean the mirrors by ion sputtering of the impurity from the reflective surface in a gas discharge. In this work, results are presented from experimental studies of the possibility of controlling the mirror cleaning process in the glow discharge localized inside a grid hollow cathode and controlled by direct current (DC) or mid-frequency pulsed direct current (MDC). Three independent methods of control were studied: measurements of the voltage of the cleaning discharge at a fixed current, measurements of the frequency of the appearance of microarc couplings in the presence of dielectric contaminants, and emission spectroscopy studies of the cleaning discharge. Separately, to verify the cleaning effect, results were presented from measuring the reflection spectrum of the mirror. It was shown that all three methods can be used in the cleaning systems. The first of the discussed methods appears to be the most promising since it does not require additional detectors and uses only the data obtained from the discharge power supply unit, and it is applicable to cleaning mirrors from both current-conducting and thin dielectric contaminations. The control based on measuring the frequency of appearance of microarc couplings can be used when one alternates between the PDC and the DC regimes in the presence of layered metal–dielectric contaminations and cleaning the local dielectric contaminations resulting from emergencies. The spectroscopic control method provides the highest sensitivity and allows one to determine the composition of the contaminants, yet it also has higher requirements for the recording equipment. The considered methods can be used to automate the processes of removing the contamination in the systems for plasma cleaning of the input diagnostic mirrors of fusion installations and to control the finishing treatment of substrates before vacuum deposition of coatings.

A comparative study of the effect of irradiation with helium ions on the optical properties of single-crystal molybdenum mirrors with crystallographic orientations 〈110〉 and 〈111〉 is presented. The irradiation mode corresponds to the conditions in the built-in systems for plasma cleaning of the input mirrors from contamination in the optical diagnostics of the ITER international thermonuclear reactor using helium as a working gas. Such irradiation results in a variation in the specular reflection and diffuse scattering of the mirror that is virtually independent of the initial structure of the surface layer of the mirror and the duration of the irradiation process. The authors explain the changes of the optical characteristics of the mirror using the formation of nanosized bubbles in the surface skin layer. The Rayleigh scattering of incident radiation by the bubbles leads to a decrease in the intensity of specular reflection and an increase in the diffuse scattering. A model of the formation and growth of nanosized bubbles and their effect on optical properties is proposed. The results must be taken into account in the analysis of experimental data in the ITER optical diagnostics after removing of contaminants using helium, when choosing a working gas for plasma cleaning systems for mirrors, and also for the formation of a nanoporous structure in a thin surface layer of metals.

The results of a study of the effects of a gas flow through the discharge region and the location of the diagnostic port walls close to the discharge cell on the efficiency of the plasma system for cleaning the entrance mirror of the Active Spectroscopy optical diagnostics, which have been developed in Russia for the ITER reactor, are presented. The experiments were carried out on a full-size functional model of a discharge cell operating on a constant or unipolar pulse current, electrically isolated from the walls of the vacuum chamber. The material of the model of the entrance mirror is molybdenum. The working gas is argon. A vacuum unit with dynamic stabilization of the working gas pressure was used. The absence of a gas flow through the discharge region and the proximity of the ITER diagnostic port walls to the discharge cell were simulated by placing the discharge cell inside a box-shaped gas-dynamic fairing placed in a vacuum chamber. Experiments with a gas flow at a large distance from the discharge cell to the walls of the vacuum chamber were carried out without using a fairing. It was revealed that the absence of a gas flow at a small distance of the discharge cell elements from the walls of the vacuum chamber increases the cleaning efficiency.

The results of a study into the limiting operating conditions for gas pressure when using D 2 , He, Ne, and Ar in the “Active spectroscopy” diagnostic plasma cleaning system for the input mirror of the ITER reactor are presented. The experiments were carried out on a full-sized functional model of a cleaning system based on a discharge with a mesh hollow cathode, with all cathode elements of the discharge cell grounded, including a full-size model of the input mirror. It was found that, when using D 2 , He, and Ne as working gases, the ignition pressure at a voltage of 750 V is approximately 60, 100, and 20 Pa, respectively, and significantly exceeds the maximum allowable for ITER in the cleaning mode (7 Pa). In this case, the discharge quenching pressure in the considered cases is ≤1 Pa. It is shown that the use of an additional thermionic cathode located inside the discharge cell makes it possible to initialize the discharge at pressures close to the quenching pressure. Turning on the hot cathode with a smooth increase in the heating current in the absence of a discharge current and turning it off after ignition of the discharge provide a significant increase in the life of the hot cathode. The applicability of all gases available for use in ITER and a radical reduction in the operating pressure facilitates the solution of the problem of degassing the vacuum chamber and diagnostic ports of thermonuclear facilities after cleaning the mirrors, which makes it possible to reduce the proportion of heavy impurities in the deuterium–tritium mixture and, accordingly, to increase the plasma temperature during operation impulse.

The results of a comparative study of the effect of irradiation with helium ions on the optical properties of single-crystal molybdenum mirrors with crystallographic orientations <110> and <111> are presented. The irradiation regime corresponds to the conditions in plasma cleaning systems of the entrance mirrors from contamination in optical diagnostics of ITER while helium is using as a working gas. As a result of such irradiation, a change in specular reflection and diffuse scattering of the mirror occurs, which is practically independent of the initial structure of the surface layer of the mirror and the duration of the irradiation process. Authors explain the observed changes in the optical characteristics of the mirror by the formation of nanosized bubbles in the surface skin layer. Rayleigh scattering of incident radiation on these bubles decreases the intensity of specular reflection and increases diffuse scattering. A model of the formation and growth of nanosized bubbles and their effect on optical properties is proposed. The obtained results should be taken into account when analyzing experimental data in optical diagnostics of ITER after removing contaminants using helium, when choosing a working gas for mirror cleaning, and also for the formation of a nanoporous structure in a thin surface layer of metals.

The results of experimental investigations of the angular distributions during magnetron sputtering of Mg, Al, Si, Ti, Cr, Cu, Zn, Ge, Zr, Nb, Mo, Ag, In, Sn, W, Pt, Au, and Bi targets in argon at a constant current are presented. The operating conditions during the experiments corresponded to the values typical of industrial sputtering equipment. The angular distribution of the material flow was calculated based on the measured thickness of the coating deposited on two flexible ribbon substrates. These substrates were fixed in place on a substrate holder in the form of two crossed half-rings that were equidistant from the center of the magnetron target (the curvature radius was 100 mm). The influence of the magnitude and shape of the magnetic field near the surface of the sputtered cathode on the angular distribution was also investigated. The results can be used as the source data for calculating the profile of the coating during magnetron sputtering.

An analytical model of the sputter deposition on an inclined flat rotating substrate, which is shifted relative to the magnetron, is developed for a magnetron with a disk cathode. The angular distribution of sputtered atoms, the profile of the cathode erosion zone, and the energy efficiency of sputtering that were measured directly on the magnetron, were used as the initial data. The magnetron was used to verify the results of calculations. This made it possible to indirectly take such effects as scattering and diffusion of the sputtered material during its transport to the substrate into account, as well as the design features of the magnetron. Thecalculation and experimental verification of its accuracy were carried out for Cu. The results of the numerical analysis of the presence of such a substrate position, which is optimal according to the criterion of the maximum sputtered-material transfer to the substrate, are presented. It was shown that there is only one set of optimal positioning parameters that meet this requirement. A comparative analysis of the efficiency of material transfer with and without a substrate tilt was performed. It was shown that a tilt provides a significant (more than 60% for the magnetron used) increase in the material transfer. These results can be used to optimize the sputter deposition process with increased requirements for the thickness uniformity of a coating and for the complex development of the deposition equipment.

Structure, optical properties, and resistance to sputtering are studied for a reflecting Mo-coating that is fabricated using magnetron deposition with simultaneous low-energy ion sputtering at the deposition rate that is higher than the etching rate. A Mo-polycrystalline mirror is used as a substrate. It is shown that the coating exhibits textured nanocrystalline structure with a relatively low spread of crystallite sizes and high resistance to sputtering. It is also demonstrated that the spectral reflection coefficient of such a Mo-coating differs from the spectral reflection coefficient of polycrystalline and single-crystalline Mo and the difference results from the effect of the structure of coating on its optical properties. A theoretical model of the coating formation is proposed.

A discharge cell that is electrically insulated from a vacuum chamber is described. The cell consists of a cylindrical mesh hollow cathode and a planar magnetron with a disk cathode, an unbalanced magnetic system, and an embedded insulated annular anode. The cell provides both cleaning of a substrate before the deposition and subsequent magnetron deposition of a coating. The magnetron is located on the side of the open end of the hollow cathode and the substrate is fixed in place inside the hollow cathode on the wall opposite to its open end. In the cleaning mode, the power supply is inverted, and the magnetron is used as the anode for the hollow cathode discharge. A discharge cell of this type provides high uniformity of cleaning, an increase in the cleaning rate, and a decrease in the minimum operating Ar pressure to ≈0.1 Pa in the substrate-sputtering mode. The operating pressure for the magnetron in the deposition mode is in the range of 0.3–1.0 Pa. The ability to rapidly transfer from the cleaning mode to the magnetron deposition by switching the power source without changing the pressure in the vacuum chamber positively affects the adhesion properties of the deposited coating.

It has been shown that the use of the survival function of the Weibull distribution shifted along the ordinate axis allows one to increase the accuracy of the approximation of the normalized profile of an erosion zone in the area from the axis to the maximum sputtering region compared with the previously suggested distribution function of the extremum values. The survival function of the Weibull distribution is used in the area from the maximum to the outer boundary of an erosion zone. The major advantage of using the new approximation is observed for magnetrons with a large central nonsputtered spot and for magnetrons with substantial sputtering in the paraxial zone.

Краткие сообщения 04 Повышение точности аппроксимации профиля зоны эрозии планарных магнетронов © А.В

The surface morphology of optical sputter-resistant nanocrystal molybdenum coatings formed by magnetron sputtering is investigated. The coatings deposited onto polished polycrystalline molybdenum mirrors have defects of two types: large crystalline inclusions protruding above the coating surface and areas with sharp boundaries corresponding to those of grains on the polycrystalline substrate surface. Preliminary treatment of the substrate surface by sputtering with the simultaneous deposition of substrate material removed defects of both types. A theoretical model of coating formation explaining the mechanism of improvement in its structural uniformity during surface preparation by the proposed method is presented.

We have proposed a universal approximation of the normalized erosion zone profile of planar magnetrons with a disk cathode by a composite function that includes the probability density function for the minimal distribution of extremal values for the region from the center of the disk cathode to the maximum of the erosion zone and the survival Weibull distribution function (from the maximum to the outer boundary of the sputtering zone). The accuracy of the approximation has been verified for six magnetrons differing in the cathode size or in the design of the magnetic systems. In all cases, good agreement has been observed between the approximation and experimentally measured values. The results reported here can be used to analyze processes that occur on the cathode during sputtering and to refine the calculations of coating profiles.

We report on the results of experimental study of the dependence of sputtering energy efficiency K (w) in a dc planar magnetron sputtering setup on the discharge power, working gas pressure, magnetic field, cathode erosion depth, and the structure of the gas puffing system and anode. We propose that this parameter be used for comparing the degree of perfection of the magnetron design irrespective of the magnetron size and structural features. The results of measurements of K (w) in sputtering of Al, Ti, Cr, Cu, Zn, Zr, Nb, Mo, Ag, In, Sn, Ta, W, Pt, and Au are considered. The optimization criterion is worked out for the magnetic system of the magnetron, which ensures the minimal working pressure and the maximal sputtering rate for the cathode. The results are analyzed theoretically.

The possibility of using the reflective gas discharge at a direct current (DC) in a strong uniform magnetic field for cleaning mirrors in optical diagnostics of the ITER is studied. A mirror and an additional Penning-type electrode, which was shifted along the magnetic field, were used as sputtered cathodes forming the discharge cell. The experiments were performed in gaseous He, in a magnetic field of up to 0.5 T at a discharge voltage of up to 750 V. Mo, Al and Mo mirrors with an Al coating, simulating a Be impurity, were used as materials of the cathodes. It is shown that this system efficiently cleans Mo mirrors from Al films with a thickness of up to 200 nm, recovering the initial optical properties of the mirror. Integration versions of cleaning systems of this type into the first-mirror assembly for the ITER are proposed for the example of Hα diagnostics.

Results of investigation of Mo polycrystalline mirrors surface modification at sputtering with mirror material deposition are presented.Surface smoothing was experimentally shown to occur if sputtering rate and proper material deposition rate are approximately equal.Computer modeling of this process shows that for the surface smoothing sputtering and deposition rates should not differ more than 2 times.